Risk-Standardized Mortality Rates as Quality Proxy for Surgical Oncology

Get Permission

In a German study reported in the Journal of Clinical Oncology, Baum et al found that risk-standardized mortality rates may be a better-quality proxy for complex oncologic surgeries than hospital volume.

As stated by the investigators, “Despite a long-known association between annual hospital volume and outcome, little progress has been made in shifting high-risk surgery to safer hospitals. This study investigates whether risk-standardized mortality rates could serve as a stronger proxy for surgical quality than volume.”

Study Details

The study included data from all patients who underwent complex oncologic surgeries in Germany between 2010 and 2018 for any of five major cancer types (esophageal, lung, pancreatic, stomach, colorectal), with the data divided into training (2010–2015) and validation sets (2016–2018). For each surgical group, annual volume and risk-standardized mortality rates quintiles were calculated in the training set, with the identified thresholds being applied to the validation set.

Risk-standardized mortality rates for each hospital were obtained as the ratio of predicted-to-expected in-hospital mortality, multiplied by the national unadjusted rate. Hospitals in the lowest quintile and highest quintile of annual caseloads were classified as low-volume and high-volume, and hospitals in the lowest quintile and highest quintile of risk-standardized mortality rates were classified as low-mortality and high-mortality; hospitals lying between the lower and upper quintiles were classified as mid-volume or mid-mortality.

Key Findings

In the validation set, 158,079 patients were treated in 974 hospitals between 2016 and 2018. With some variation dependent on type of surgery, ≥ 50% of high-volume hospitals were not ranked within the low-mortality hospitals according to risk-standardized mortality rate grouping. For example, for lung cancer surgery, 57.1% and 14.3% of high-volume hospitals were in the medium- and high-mortality risk-standardized mortality rate groups. For stomach cancer surgery, 42.1% and 21.1% of high-volume hospitals were in the medium- and high-mortality risk-standardized mortality rate groups. For colorectal cancer surgeries, 58.2% and 10.4% of high-volume hospitals were in the medium- and high-mortality risk-standardized mortality rate groups.  

In evaluation of the effectiveness and efficiency of risk-standardized mortality rate–based and volume-based approaches to moving patients away from low-performing hospitals, it was found that risk-standardized mortality rate centralization would have resulted in a total of 955 lives saved across all surgery types between 2016 and 2018, compared with 663 lives saved with volume-based centralization. For the risk-standardized mortality rate–based approach, 30,253 patients would have to be moved, yielding an average of 32 patients being relocated to a low-mortality hospital to save one life. For the volume-based approach, 31,155 patients would have to be moved to a high-volume hospital, yielding an average 42 being relocated to save one life.

Use of risk-standardized mortality rate–based centralization was estimated to significantly reduce (all P < .001) median patient travel time vs volume-based centralization for each of the surgeries considered (esophageal = 17 vs 27 minutes, lung = 15 vs 26 minutes, stomach = 9 vs 16 minutes, pancreatic = 13 vs 23 minutes, and colorectal = 9 vs 14 minutes).

The investigators concluded: “Risk-standardized mortality rate is a promising proxy for measuring surgical quality. It outperforms volume in effectiveness, efficiency, and hospital availability for patients.”

Philip Baum, MD, of the Department of Thoracic Surgery, Thoraxklinik at Heidelberg University Hospital, is the corresponding author for the Journal of Clinical Oncology article.

Disclosure: For full disclosures of the study authors, visit

The content in this post has not been reviewed by the American Society of Clinical Oncology, Inc. (ASCO®) and does not necessarily reflect the ideas and opinions of ASCO®.